When the population of an endangered marine mollusc (Patella ferruginea) increases almost three-fold in ten years. Reality or fiction?

When the population of an endangered marine mollusc (Patella ferruginea) increases almost three-fold in ten years. Reality or fiction? E. Ostale ´-Valriberas*,A ´ . Sabino-Lorenzo, A. Ali-Ahmed, A. Pavo ´n-Paneque, J. Sempere-Valverde, F. Espinosa and J. C. Garcı ´a-Go ´mez Laboratorio de Biologı ´a Marina de la Universidad de Sevilla (LBMUS)/ A ´rea de Investigacio ´n I+D+i del Acuario de Sevilla/ Estacio ´n de Biologı ´a Marina del Estrecho (Ceuta), Universidad de Sevilla, Seville, Spain The critically endangered species Patella ferruginea (Gastropoda, Patellidae), endemic to the western Mediterranean, has breeding populations in both natural and artificial habitats, the latter of which are generally linked to port infrastructures. Over the past decade, the temporal change of this species’ population has been monitored (structure and density) using exhaustive censuses along Ceuta’s coast (Strait of Gibraltar), one of the few stronghold populations within the entire Mediterranean basin. This study focuses on the population dynamics of P. ferruginea in Ceuta and the environmental factors that affect the structure of this population, such as wave exposure, coastline heterogeneity, substratum roughness, substratum lithology, and chlorophyll-a concentration. Different potential negative interactions were also considered: angling, shell fishing, bathing in the intertidal, bathing near the intertidal, recreational boating and temporary migrant campsites nearby. The results have shown in the period 2011-2021, the estimated size of P. ferruginea population has increased by 200 %, from 55,902 to 168,463 individuals (of which 131,776 are adults). The subpopulation with the greatest increase in these years was the one settled on dolomitic rip-raps inside the Ceuta’s harbor, with an increase of 1,288%. The results of the present study indicate that Ceuta hosts the main population of this endangered species through its distributional range (Western Mediterranean), being a source population on the Southern Iberian Peninsula that its preservation must be prioritized. Statistical modelling has shown that the adult density of P. ferruginea is positively influenced by coastal heterogeneity, habitat area and substratum roughness, but negatively by vertical inclination, concentration of chlorophyll-a, and anthropogenic impact. These results also support the concept of ¨Artificial Marine Micro-Reserves¨ as a new area-based conservation measure according with the IUCN guidelines, as these will contribute to setting up a network of these source populations that promote genetic flow among populations, with eventual recolonization throughout its original distribution. KEYWORDS Patella ferruginea, patellogastropoda, conservation, extinction risk, population structure, artificial marine micro-reserves, artificial substratum, source population Frontiers in Marine Science frontiersin.org01 OPEN ACCESS EDITED BY Alan Hodgson, Rhodes University, South Africa REVIEWED BY Francesco Tiralongo, University of Catania, Italy Mariachiara Chiantore, University of Genoa, Italy *CORRESPONDENCE E. Ostale ´-Valriberas [email protected] RECEIVED 19 December 2022 ACCEPTED 08 June 2023 PUBLISHED 12 July 2023 CITATION Ostale ´-Valriberas E, Sabino-Lorenzo A ´, Ali-Ahmed A, Pavo ´n-Paneque A, Sempere-Valverde J, Espinosa F and Garcı ´a-Go ´mez JC (2023) When the population of an endangered marine mollusc (Patella ferruginea) increases almost three-fold in ten years. Reality or fiction? Front. Mar. Sci. 10:1127630. doi: 10.3389/fmars.2023.1127630 COPYRIGHT ©2023Ostale ´-Valriberas, Sabino-Lorenzo, Ali-Ahmed, Pavo ´n-Paneque, Sempere-Valverde, Espinosa and Garcı ´a-Go ´mez.Thisisanopen-access article distributed under the terms of the Creative Commons Attribution License (CC BY). The use, distribution or reproduction in other forums is permitted, provided the original author(s) and the copyright owner(s) are credited and that the original publication in this journal is cited, in accordance with accepted academic practice. No use, distribution or reproduction is permitted which does not comply with these terms. TYPE Original Research PUBLISHED 12 July 2023 DOI 10.3389/fmars.2023.1127630 1 Introduction Coastal areas host half of the world’s cities and more than half of the world’s population, with population density near the coast three times higher than the global average (Creel, 2003;Small and Nicholls, 2003;Airoldi and Beck, 2007). Shoreline urbanization combined with the increase of tourist, recreational and commercial activities, has resulted in the introduction and proliferation of artificial structures in marine coastal habitats worldwide (Bulleri and Chapman, 2010;Firth et al., 2016). The human impact associated with these structures has been reported to affect biological processes on all spatial and temporal scales (Airoldi et al., 2005). Among them, fragmentation of natural habitats by urbanization processes can cause a reduction in population size and connectivity of species, and thus increase their risk of extinction in the face of certain permanent disturbances (Aguilera et al., 2020; Chase et al., 2020). Limpets are considered keystone species in rocky coastal ecosystems as they maintain the composition and structure of the biological community and, when they disappear from their habitats, cause a cascading effect on them (Raffaelli and Hawkins, 1996). Therefore, limpets may be the most important group of grazers in temperate rocky intertidal zones, because their feeding activity limits the growth of macroalgae and controls barnacle numbers (Branch, 1981). Among limpets, the larger species are frequently the most ecologically relevant, being defined as ‘giant limpets’(shell length greater than 100 mm). With this classification a total of 14 species in the world can be found, three of them catalogued as threatened (Espinosa and Rivera-Ingraham, 2017). Patella ferruginea Gmelin, 1791 (Gastropoda: Patellidae) belongs to this classification, as it achieves a maximum recorded shell length of 115 mm (Supplementary Figure 1-S). It is endemic to the Western Mediterranean Sea and it is considered the most endangered marine invertebrate (Ramos, 1998), together with Pinna nobilis,withinthe European coastline (Vazquez-Luis et al., 2017;Lattos et al., 2021). Like other limpet species, P. ferruginea is a broadcast spawner with external fertilization so that the urban development of coastal areas has dramatically fragmented its habitat (Airoldi et al., 2005; Firth et al., 2016). Moreover, the conspicuous star‐shaped shell and large sizes that P. ferruginea can reach has made this species the target of human collection since the Quaternary (Espinosa, 2006; Espinosa et al., 2009;Rivera-Ingraham et al., 2011a;Guallart et al., 2013a). On account of these anthropogenic impacts, its distribution range has progressively been contracted, and P. ferruginea has nearly disappeared from European continental coasts, except for the small and fragmented populations present in Andalusia Southern Spain, Liguria (Italy) and Southern France (Cottalorda et al., 2004;Espinosa et al., 2014b;Luque et al., 2018;Ferranti et al., 2019;CMAOT, 2020). The remaining most notable populations are found in scattered North African gated communities: Ceuta, Melilla, Chafarinas Islands, Habibas Islands, and the island of Zembra (Espinosa et al., 2014b;Luque et al., 2018). In this regard, it has recently been identified that possibly the population located in Ceuta (Strait of Gibraltar, North Africa) plays the role of source population in the area, maintaining nearby sink populations in the south of the Iberian Peninsula (Espinosa et al., 2018) according to the “source-sink population model”(see Pulliam, 1988; Pulliam, 1996). This study focuses on the population dynamics of P. ferruginea in Ceuta and the environmental factors that shape the population size structure. Consequently, the objectives of the present study are i) to determine the current status of this P. ferruginea population, ii) track the any change in the population in the last 10 years, iii) to confirm the result of the ¨population viability analysis¨ (PVA) prediction method performed by Espinosa et al. (2018) of the P. ferruginea population in Ceuta and iv) explore the environmental factors affecting the populations of P. ferruginea. 2 Materials and methods 2.1 Study area The Strait of Gibraltar is a very important geological region, borderingbytheMediterranean, Lusitanian and Mauritanian biogeographic provinces, so it is a zone in which marine flora and fauna from the Mediterranean and Atlantic overlap (Coll et al., 2010). Likewise, it is one of the world’s busiest hubs for maritime traffic, as it is a place where many commercial routes converge and where important harbors are present (Algeciras, Gibraltar and Tangier Med), which makes it an area with high risk of disturbances, impacts and environmental disasters (Piniella and Walliser, 2013:Nachite et al., 2020). The harbor of Ceuta, a Spanish city in North Africa, has an unusual structure, because it is located between two bays (North and South) connected to both by the Royal Moat (Figure 1A). This defensive construction, built by the Portuguese during the 16 th century, increases water movement and exchange within the structure (shorting the residencetimeofthewatermass),whencomparedtoother conventional harbors, maintaining moderate oxygen levels in the water and sediment (Guerra-Garcı  a et al., 2004a;Guerra-Garcia and Garcı  a-Gomez, 2005;Sanchez-Badorrey et al., 2021). 2.2 Sampling methods 2.2.1 Patella ferruginea census A census on P. ferruginea population in Ceuta was carried out during 2019 and 2020 (14 months). To estimate the status and evolution of the P. ferruginea population, the sampling methodology repeated the one carried out by Rivera-Ingraham et al. (2011b) in 2010 in the same area. The complete coastline was divided into 18 sectors; accounting for the type of substratum (natural, dolomitic rip-rap and concrete cubes) and its geographic orientation (see sectors at Figure 1B). To know the ¨population size structure¨ of P. ferruginea’s population, the 20% of the entire rocky midlittoral, where the species occurs, was sampled at each sector. Ten meter linear transects were proportionally distributed within each sector separated by 40 m. The shell length of the specimens found Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org02 within the intertidal at each transect were measured to the nearest millimetre using a calliper (Guerra-Garcı  a et al., 2004b;Espinosa, 2009b;Rivera-Ingraham et al., 2011b). Finally, 4,560 m of rocky midlittoral coastline were sampled out of a total of 23,617 m of rocky coastline present in the city of Ceuta (Table 1). To estimate the total number of individuals in each sector and thus of the population, the results were extrapolated from the total length (m -1 )ofP. ferruginea habitat in the sector. In the present study, recruits were defined as individuals smaller than 30 mm and adults greater than or equal to 30 mm (Espinosa et al., 2006) and the “class size”was categorize for every 10 mm of shell length (e.g., class size 3 pertains to individuals from 21mm to 30mm). To track any change of the population structure in the last 10 years, it has been considered that in the preceding work performed by Rivera-Ingraham et al. (2011b). In the said study, the total length of rocky shoreline where the species could potentially be present was calculated by using 1:9,000 scale maps obtained using Google Earth © . Therefore, the estimations from the previous census have been recalculated, according to the used estimate of coastline length in the present work (10 m resolution), in order to make both estimations comparable and to be able to evaluate the percent increase or decrease of the population in each sector. The percentage change in the number of individuals (%) between both studies was calculated using the equation: Percentage change (%)=(estimation  2020  estimation  2010  )x 100) −100 2.2.2 Environmental factors Once the sampling of P. ferruginea was completed, the environmental factors were studied. All environmental factors were studied in the same transect that were randomly selected in each sector (three transects per sector, 54 transects in total) in July of 2022. Substratum lithology of each sector was characterized by collecting three rock chips from each sector, which were powdered in the laboratory using a ball mill with stainless steel balls as grinding media. Each rock chip was obtained by hammer and chisel. The elemental composition and calcination percentage of each sample was quantified by X- ray fluorescence (XRF) using an AXIOS spectrometer. Mineralogic absorption spectra were detected by X- ray diffraction (XRD) using a powder diffractometer (Bruker D8 Advance) equipped with a high temperature chamber (Anton Paar XRK 900) and a fast response/high sensitivity detector (Bruker Vantec 1) with radial Soller slits (full methodology can be accessed in Valverde et al., 2015). Afterwards, XRF and XRD data were interpreted to quantify the crystallinity and lithology composition of each sample using DRIFAC.EVA.4.1 software. To assess the influence of microalgal abundance on P. ferruginea demographic, the chlorophyll-a concentration on the substratum surface was estimated in the same rock samples used to study lithology. The resulting slurry was treated with 20 ml of 100% ethanol to extract the pigments. Then, samples were frozen and maintained in darkness until analysis (minimum of 48 h) according with the protocol used by Kido and Murray (2003). Samples were filtered through a Whatman GF/C filter and measurements of absorbance were carried out with a spectrophotometer (Pharmacia Biotech Novaspec II). Given the large extension of the study area, some environmental factors like chlorophyll a were collected once, being a snapshot of the environment. Moreover, for the Chl-a we took the 54 samples in two consecutive days of July 2020 using a zodiac and sampling in FIGURE 1 (A) Location of Ceuta, inside the black circle can be found the royal Moat; (B) Sectorization of the coastline to estimate the population size distribution of P. ferruginea. Source: Google Earth. Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org03 TABLE 1 Comparison of the mean density, adult mean density and the total estimate individuals between both studies about P. ferruginea population of Ceuta (change from 2010 to 2020). Sector Rivera-Ingramham et al., 2011b Present study Percentage change in the estimated number of the limpet, present study respect to Rivera-Ingramham et al., 2011b Total habitat length (m) Sampled length (m) Total individuals registered (ind.) Density (ind./m) Adult mean density (ind./m) Total estimate individuals Total estimate individuals according the total habitat length of the present study Total habitat length (m) Sampled length (m) Total individuals registered (Ind.) Density (ind./m) Adult mean density (ind./m) Total estimate individuals A 300 100 195 1.95 1.72 585 1,043.25 535 110 115 1.05 0.87 559.32 -46.39 B 311 311 2,260 7.27 5.83 2,260 5,922.51 815 170 672 3.95 3.38 3,221.65 -45.60 C 712 712 3,518 4.94 3.25 3,518 3,952.81 800 160 808 5.05 3.88 4,040.00 2.21 D 350 350 2,262 6.46 4.28 2,262 4,200.86 650 130 788 6.06 5.30 3,940.00 -6.21 E 1142 100 104 1.04 1.03 1,187 1,352.00 1,300 260 1,286 4.95 4.68 6,430.00 375.59 F 350 100 106 1.06 0.86 371 848.00 800 160 874 5.46 5.03 4,370.00 415.33 G 1,710 100 297 2.97 2.22 5,079 6,237.00 2,100 420 2,126 5.06 4.39 10,630.00 70.43 H 2,201 258 479 1.86 1.79 4,086 3,434.69 1,850 370 3,263 8.82 7.89 16,315.00 375.01 I 1,271 100 440 4.40 3.56 5,592 10,340.00 2,350 470 4,921 10.47 8.81 24,605.00 137.96 J 959 100 19 0.19 0.15 182 313.50 1,650 330 277 0.84 0.77 1,385.00 341.79 K 555 100 709 7.09 4.28 3,934 4,254.00 600 120 2,435 20.29 15.35 12,175.00 186.20 L+L’3,754 3,754 4,111 1.10 0.66 4,111 2,792.50 2,550 510 7,755 15,21 9.27 38,775.00 1,288.54 M 1,115 50 244 4.88 3.70 5,441 5,616.88 1,151 230 3,065 13.33 9.01 15,338.33 173.08 N 1,156 30 23 0.77 0.10 886 881.67 1,150 230 1,536 6.68 4.79 7,680.00 771.08 O 316 100 273 2.73 1.71 863 682.50 250 50 661 13.22 9.04 3,305.00 384.25 P 1,217 163 351 2.15 0.21 2,987 2,963.04 1,376 270 1,964 7.27 5.21 10,009.13 237.80 Q 1,139 100 42 0.42 0.37 477 667.80 1,590 320 785 2.45 2.00 3,900.47 484.08 R 900 100 19 0.19 0.18 171 399.00 2,100 420 357 0.85 0.76 1,785.00 347.37 Total 19,458 6,628 15,452 2.86 1.99 43,992 55,902.01 23,617 4,560 33,687 7.28 5.58 168,463.89 201.36 Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org04 the same hour of the day. The chlorophyll-a concentration was calculated according to the formula of Thompson et al. (1999): ½Chlorophyll −a=12:2xA665  xv (dxV) Where, 12.2 = constant for ethanol, A6655 = net absorbance of solution at 665 nm, v = final volume of solution (20 ml), d = path length of cell (1.6 cm), V = surface area of sample (100cm 2 ). Chlorophyll-a concentration is expressed in μg cm -2 . The wave exposure was quantifiedforeachsectorbasedonafetch model index developed by Howes et al. (1994) (see Terron-Sigler et al., 2018). Fetch models have been successfully used to predict marine community patterns (e.g. Ros et al., 2016)byproviding quantitative estimates of wave exposure using a combination of two indices: maximum fetch (Fi) and modified effective fetch (Fe). Maximum fetch is defined as the maximum fetch distance in km measured from the point of interest. When a vector does not find and obstacle (i.e. open ocean), a value of 1,000 km is conventionally used. Effective fetch (Fe) is calculated from the equation: Fe =½o(cos Өi)Fi ocos Өi:Fe Where Өi is the angle between the shore-normal and the directions 0°, 45° to the left ad 45° to the right, and Fi is the fetch distance in km along the relevant vector. Combining the values obtained for each index, the wave exposure class of each sector was determined based on the classification proposed by Howes et al. (1994). The different Fi values (-45°, 0° and 45°) were calculated for each transect studied using Google Earth © . The heterogeneity of the coast (macroscale roughness) was calculated by lying a 10 m flexible measuring tape along the upper mid-littoral, which was allocated following as closely as possible the contours of the bare substratum along the transect. Using the same methodology, three replicates were recorded on each transect randomly selected along the height of the rocky midlittoral (habitat of P. ferruginea). This information was used to obtain was used to obtain the area of P. ferruginea habitat in each sector: profile of longitudinal x profile of height (m). Regarding microscale roughness, three 25 cm profile gauges per transect with 0.5 mm pins were pushed onto the bare rock within the same transects already selected for macroscale roughness to record substrata surface heterogeneity (Frost et al., 2005). The resulting profiles were photographed, and the images were digitally processed with Adobe © Photoshop to obtain twocoloured images. The length of the contour of the profile was obtained with ImageJ software. In both cases (macro- and microscale), substratum roughness was calculated as in Rivera- Ingraham et al., 2011b;Rivera-Ingraham et al., 2011c and Sedano et al., 2020 using the equation by Blanchard and Bourget (1999): Roughness =Tr Ts. Where Tr is the contour measured between two points and Ts the linear distance between those points. The inclination of the substratum was taken recording three replicates within the already selected transects using a digital clinometer (see Rivera-Ingraham et al., 2011c). The impact of human activity in the limpet’s habitat (intertidal zone) was sampled in July 2021, since it is the month of the year when more people are present within the littoral (personal observations from Ostale-Valriberas). In the present study, the variable “impact of human activity”was restricted to the direct interaction of the citizen with the rocky intertidal. Each sector was visited twice by boat to quantify the different interactions. Sampling was conducted on two weekends in July due to the large number of citizens around the shoreline on these days. The first weekend sampled was from 11 a.m. to 2 p.m. in the morning and the second from 6 p.m. to 9 p.m., each sector was observed for half an hour each day. In each sampling session, all the sectors were visited. Each interaction was scored from 1 to 3 to categorise the different impacts on P. ferruginea habitat. The difference interactions registered were angling (x3), shell fishing (x3), bathing in the intertidal (x2), bathing near the intertidal (x1), recreational boat nearby (x1) and irregular camp (x1) (the concept ¨irregular immigrant camp¨ derives from the exceptional migratory problem of 17th May of 2021 in which the local administration assistance was overwhelmed, see investigation of Casey and Bautista (2021). To estimate the used index for the GLM, the total of each interaction was multiplied by the impact value on the habitat (e.g. 9 people bathing in the intertidal (x2) gives a value of 18). In the methodology for quantifying human activity, value 1 is “low impact” and 3 is “high impact”. A constrained ordination approach to assess how well the biological data relate to different abiotic variables that characterize the different sectors was carried out. The parameters considered were heterogeneity of the coast (macroscale roughness), microscale roughness, area of P. ferruginea habitat, concentration of chlorophyll-a, inclination of the substratum, wave exposure (fetch) of the littoral, impact of human activity and rock elemental composition. 2.3 Statistical analyses 2.3.1 Patella ferruginea census Among the sectors, a PERMANOVA analysis was carried out between the different size classes to differentiate the population size structure of P. ferruginea distribution. Before the analyses, a square root transformation along with the Bray Curtis similarity index was applied to different size classes. Differences among sectors were established using nNMDS (non-metric multidimensional scaling) and cluster analysis. The change of the P. ferruginea population in the last 10 years was determined by comparing the adult density between the present study and preceding work performed by Rivera-Ingraham et al. (2011b). To obtain this result, Euclidean distance matrices were calculated and tested using univariate PERMANOVA. PERMANOVA analyses were performed using the PRIMER v.6 +PERMANOVA package (Clarke and Gorley, 2005;Anderson et al., 2008). 2.3.2 Environmental parameters The effect in the P. ferruginea population’s structure (adult density and adult mean size) of different environmental parameters were modelled using generalized linear models (GLMs; McCullagh and Nelder, 1989). Models were carried out using normal Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org05 distributions and identity link functions to determine which of the parameters gave the best fit to the data (see Cayuela, 2010). The environmental parameters added to the model were the percentage of calcium oxide in the substratum (CaO %), chlorophyll-a concentration, wave exposure (fetch), roughness of substratum (macroscale and microscale), area of P. ferruginea habitat, inclination of the littoral and anthropogenic impact. Prior to the analysis, environmental variables were tested for multicollinearity, assuming collinearity when variance inflation factor (VIF) values were much greater than 1. Model selection was carried out based on Akaike’s information criterion for small sample sizes (AICc) (Burnham and Anderson, 2002). Variability explained by the best models was computed using deviance values (D 2 ), comparing residual deviance with the deviance of a null model (null deviance). Among all significant models, the minimal Akaike value indicated the best model that explained the greatest proportion of variance under the restriction that all the predictors in the model must be significant (see Cayuela, 2010;Ros et al., 2015). The density of recruits (individuals< 30 mm) was not considered in the GLM because their presence is highly fluctuating throughout the year (recruits of this species settle on the intertidal rock between December and March) (see Espinosa, 2006), and it was impossible to sample all sectors at the same time in those months. Generalized Lineal Models were carried out using SPSS 25.0 © (IBM, New York, NY, USA) (Pallant, 2020). 3 Results 3.1 Present status of Patella ferruginea in Ceuta In the 20% of the P. ferruginea habitat (4,560 m), it was registered 33,687 individuals so that it was estimate that the population of Ceuta have 168,464 individuals distributed in a total of 23,617 m of rocky coastline (Table 1). From the whole population, it was estimated that 131,775.87 were adult individuals (≥30 mm). PERMANOVA analysis showed that there were significant differences between the population structure (different size classes) of the different sectors (Table 2). Individuals of P. ferruginea were present in all sampled sectors (Figure 2), with the greater mean density of individuals and adults occurring in the dolomitic rip-rap inside the harbour (sector L), with a mean density of 32.2 ind./m and adult mean density of 19.57 ind/m (Figure 3). The greatest values per transect were achieved in this sector (54.84 ind./m and 30.84 ind./m of adults), although the concrete seawalls in the same area (subsector L’) had one of the lowest values. The natural habitat with greater P. ferruginea TABLE 2 One-way PERMANOVA results for P. ferruginea size classes on different studied sectors. Source of variation Df MS Pseudo-F P (perm) Size clases (cm) Sector 18 10100 15.93 < 0,0001 Transect (sector) 452 633.98 Total 470 FIGURE 2 General distribution of P. ferruginea in Ceuta. Circle diameter corresponds to the density of individuals. Google Earth. Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org06 densities was Acantilados de la Sirena (sector H) with 8.82 ind./m and 7.89 ind./m of adults. Greater average shell length was registered in sector E, in which mean size was 584 mm and the adult mean size 603 mm (Figure 4). The biggest individual measured had a maximum shell length of 113 mm (Supplementary Figure 1-S). The average shell size of Ceuta’s population was 447 mm and the mean adult size 508 mm. Multivariate analyses revealed differences in the structure of P. ferruginea population (abundances within size classes) among sectors (Figures 5–7). The combination of NMDS and Cluster analyses showed the segregation of sector L from other sectors. The subpopulations most similar to sector L were K, M and O, all these sectors were composed of rip-raps made with quarry rocks (Supplementary Table 4-S). On the contrary, coinciding with the less dense sectors, were the sectors R, J, A and L’(Table 1;Figures 2,3). FIGURE 3 Mean density and adult mean density of P. ferruginea for each of the sectors and standard deviation. FIGURE 4 Mean size and adult mean size of P. ferruginea for each of the sectors and standard deviation. Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org07 3.2 Change in the population structure of the Patella ferruginea population in the last 10 years Recalculating the total estimated individuals from the study of Rivera-Ingraham et al. (2011b) according to the total habitat length of the present study, the population has increased from 55,902 to 168,463 individuals. This means that the population has increased by an average of 201.36% compared to the existing population 10 years ago (Table 1). The subpopulation with the higher increase in these years was the one inhabiting dolomitic rip-rap inside the harbor (sector L), with an increase of 1,288%. On the opposite side, the subpopulations with higher decrease were those on sectors A and B with a -46.39% and -45.60% respectively (Table 1). Univariate PERMANOVA showed that the adult mean density (ind./m) of the present study was significantly higher than the Rivera-Ingraham et al. (2011b) study (Pseudo-F = 14.68; p< 0.001). 3.3 Effects of environmental parameters on population structure According to an omnibus test, only the variable adult density was significant. The VIF values show that the variables used to evaluate their influence on adult density were not correlated with each other. FIGURE 5 nMDS computed using the abundances of each size class and showing PCO centroids for each of the sectors. Red circle: sectors with dolomitic riprap of the commercial harbor where the access is restricted with a great heterogeneity substratum (macro and micro scale). FIGURE 6 Cluster analysis computed using the mean abundances of the different size classes of P. ferruginea for each of the sectors considered. Discontinuous lines indicate significantly different groups of sectors segregated by SIMPROF analysis. Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org08 The model demonstrated that the adult density is significantly influenced by heterogeneity (macroscale), microscale roughness, inclination of the littoral, chlorophyll-a concentration and anthropogenic impact. According to AICc values, the best model was the combination of these five parameters, reporting the 85.9% of the variability explained (Table 3). Adult density showed a positive correlation with the macroscale and microscale variables, while the variables inclination, chlorophyll-a concentration and anthropogenic impact showed a negative correlation. Analyzing the results obtained for the microalgae (Chl-a concentration) in each sector, it is observed that the highest values (marked in green in Table 1-S) correspond to the natural TABLE 3 Model selection results explaining the environmental parameters that have influence in the adult density and adult mean size. df AICc bExp(b)Di LogLik D²(%) Adult density Normal distribution Link function: Logit Macro* + Micro* + Incli** + Chl* + Anthro** 12 38.82 2.35/0.24/-0.06/ -0.8/-0.38 10.52/1.27/0.94/ 0.45/0.962 0 -6.81 85.9 Macro** + Incli** + Chla* + Anthro** 13 39.08 3/-0.75/-0.66/-0.38 20.24/0.93/0.51/0.96 0.26 -9.72 80.5 Only significant models according to an omnibus test are shown. Model averaged coefficients (b) of explanatory variables in the significant models; Exp(b): exponential. Akaike’s information criterion corrected for small sample size (AICc) and AICc difference between the AICc of each model and the AICc of the best fitted model (Di) were used for comparison. Variability explained by the best models is shown using deviance (D²). The model with the lowest Di is the best AICc model. Variables considered were: Macro, macroscale; Micro, microscale; Area, surface; Incli, inclination; Chl-a, concentration of chlorophyll-a; Anthro, anthropogenic impact; Fetch, wave exposure; CaO, calcium oxide. **P< 0.001. *P< 0.05. B CD EF GH IJ KL MN OP QR A FIGURE 7 Size class frequencies of P. ferruginea for each of the sectors considered to know the population structure (trend line). Letters indicate the sampled sectors (see Figure 1B). Ostale ´-Valriberas et al. 10.3389/fmars.2023.1127630 Frontiers in Marine Science frontiersin.org09